Carbonyl chemistry: additions and eliminations

The C=O group is more reactive than a C=C by two orders of magnitude, and it reacts by an almost opposite logic. Where an alkene is electron-rich and attacked by electrophiles, a carbonyl is electron-poor at carbon and attacked by nucleophiles. This guide sorts the whole carbonyl family — aldehydes, ketones, acids, esters, amides, acyl chlorides, and anhydrides — onto one reactivity ladder, then walks the two mechanistic families (addition and addition-elimination) that account for almost every reaction the group does.

Prerequisites: electrophilic addition to alkenes sets the alkene baseline these mechanisms invert. Aromaticity matters when the carbonyl is part of an aromatic ring.

The C=O group, honestly

C=O is polarised: oxygen’s electronegativity (3.44) exceeds carbon’s (2.55), so the π bond is drawn toward oxygen. Two consequences:

1. Carbon is δ+ — a nucleophile attacks here.
2. Oxygen is δ− and holds a lone pair that can pick up a proton (or a Lewis acid).

The dipole moment of a simple aldehyde is 2.7 D (acetaldehyde, IUPAC Gold Book) versus 1.7 D for an ether — the C=O polarisation is real and reactivity-determining. When a nucleophile attacks carbon, the π electrons of the C=O flow onto oxygen, converting it from sp² carbonyl to sp³ tetrahedral intermediate with an oxyanion (or an OH after protonation).

Formally, the tetrahedral intermediate is the branch point. It can:

• Reprotonate and stay — giving an addition product. This is what aldehydes and ketones do.
• Kick out a leaving group and reform the C=O — this is addition-elimination, giving a substitution product. This is what acids and their derivatives do.

Which fate the intermediate takes is determined by whether there is a leaving group already on the carbonyl carbon.

The reactivity ladder

The carbonyl family sits on one ladder, most reactive at top:

Two variables drive the ordering:

Electrophilicity of the carbonyl carbon. More electronegative or better π-donating groups on the carbonyl reduce electrophilicity. Cl (electron-withdrawing by induction, weak π donor) leaves the C+ most exposed; NR₂ (moderate σ-withdrawer, strong π-donor) shields it.

Leaving-group ability. Once the tetrahedral intermediate forms, the group that can leave determines whether substitution happens at all. Cl⁻ is a great leaving group; R⁻ is a terrible one. Aldehydes and ketones can’t substitute because H⁻ and R⁻ don’t leave — they can only add.

The ladder is one-directional: a more reactive derivative can be converted into a less reactive one (acyl chloride → ester is easy), but the reverse requires special reagents (ester → acyl chloride needs SOCl₂ or oxalyl chloride after hydrolysis).

Aldehyde and ketone additions

The generic addition of a nucleophile Nu to a C=O:

1. Nu attacks carbonyl carbon; π electrons move to oxygen.
2. Tetrahedral intermediate (alkoxide) protonates to give the alcohol.

Some standard Nu-C=O combinations:

• Hydride (NaBH₄ / LiAlH₄) → primary alcohol from aldehyde, secondary alcohol from ketone. NaBH₄ works on both; LiAlH₄ also reduces esters and acids that NaBH₄ won’t touch.
• Grignard reagents (RMgX) → primary alcohol from formaldehyde (1 C added), secondary from other aldehydes, tertiary from ketones.
• Cyanide (NaCN, HCl) → cyanohydrin. Reversible; equilibrium favours products for aldehydes and small ketones.
• Alcohols (in acid) → hemiacetal, then acetal. This is the mechanism of sugar-ring formation.
• Amines (in mild acid) → hemiaminal → imine (primary amine) or enamine (secondary amine).

The hemiacetal-to-acetal step is not just addition — it is addition-elimination of water, giving a stable acetal that survives base but hydrolyses in acid. Acetals are the standard protecting group for aldehydes and ketones.

Worked example — imine formation

Acyl derivatives — addition-elimination

For a carbonyl with a leaving group X (Cl, OR’, O-COR, NR’₂), the tetrahedral intermediate has a choice: kick out Nu (reverse the addition) or kick out X (net substitution). If X leaves more easily than Nu, the substitution product forms:

1. Nu attacks the carbonyl carbon.
2. Tetrahedral intermediate.
3. X leaves, C=O reforms.
4. Product: R-CO-Nu.

This “addition-elimination” mechanism (a.k.a. AAC2 for esters in acid, BAC2 in base — Clayden 2e §12) governs every reaction that converts one acyl derivative into another:

• Acyl chloride + alcohol → ester (fast, RT)
• Anhydride + amine → amide (fast, RT)
• Ester + water (H⁺ or OH⁻) → carboxylic acid (slow, needs heat)
• Amide + water (H⁺ or OH⁻, hot, long time) → carboxylic acid

The relative rates track the reactivity ladder: acyl chloride reacts 10¹¹ times faster than amide with the same amine (Carey & Sundberg 5e §7.7).

Worked example — saponification of an ester

Enolate chemistry — the α-carbon side

The other half of carbonyl reactivity happens at the α-carbon (the one adjacent to C=O). C=O withdraws electron density σ-inductively, weakening the α C-H bond: pKa ≈ 20 for a simple ketone α-H, compared with pKa ≈ 45 for a normal alkane C-H (Clayden 2e §21). Deprotonate an α-H and you get an enolate — a nucleophile in its own right.

Enolate chemistry drives:

• Aldol reactions — enolate of one carbonyl attacks another carbonyl.
• Claisen condensations — the same but with esters.
• α-halogenation, α-alkylation — enolate reacts with electrophile.
• Michael additions — enolate adds to an α,β-unsaturated carbonyl.

The α-carbon is thus the enolate’s electrophile-attacking end; the C=O carbon is the nucleophile-attracting end. Every carbonyl compound has two reactive sites, and mastering both is the leap from “sophomore organic” to synthesis.

Common pitfalls

Choosing the wrong nucleophile for the wrong derivative. Grignards react with any carbonyl (kill acyl chlorides, ketones, esters — sometimes overshooting to the tertiary alcohol from an ester through a 2× addition). NaBH₄ reduces aldehydes and ketones selectively, leaving esters and amides untouched.

Ignoring the pH sensitivity of imine formation. Too acidic: the amine is protonated, no nucleophile. Too basic: the hemiaminal OH is not protonated, no leaving group. The window is ~pH 4-6.

Missing the tetrahedral intermediate. Every substitution at a carbonyl carbon proceeds through a tetrahedral intermediate; there is no SN2 backside attack at a sp² carbon. Drawing the substitution as a one-step process misses the mechanistic reality and predicts wrong stereochemistry when it matters.

Forgetting the irreversibility of saponification. Acid-catalysed ester hydrolysis is reversible (Fischer esterification runs both directions); base-catalysed hydrolysis is not, because the final carboxylate is thermodynamically trapped. Use base if you want the reaction to go to completion; use acid + excess water for equilibrium control.

Practice

Where this leads

Carbonyl chemistry is the launch pad of the second half of an organic course:

• Retrosynthesis — target-oriented planning almost always includes a carbonyl disconnection.
• Enolate chemistry — aldol, Claisen, Michael, alpha-halogenation, Robinson annulation.
• Named reactions — Wittig, Baeyer-Villiger, Beckmann, Cannizzaro; all extensions of the same addition or addition-elimination pattern.
• Bioorganic carbonyls — peptide bonds are amides; sugars are polyhemiacetals; every step of the citric acid cycle uses at least one enol/enolate transformation.

Every one of them starts by drawing the tetrahedral intermediate and asking what leaves.